Announcement

Collapse
No announcement yet.

Lens Correction Feature

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Lens Correction Feature

    Just wondering, would it be nice if ACDSee were to implement an in-application lens profiling feature. Right now it relies on a third party lens correction database, which isn't up-to-date and not very helpful if one owns a lens that isn't in the database, but couldn't ACDSee develop an algorith to reverse engineer the in-camera corrections? It could be as simple as providing 2 images per Focal Length, one in-camera JPEG and the RAW version of the sames, and compare the two. It could even use the embedded JPEG preview in the RAW file as the base for comparison. The idea seem so simple that one has to wonder why they aren't doing it already by default instead of relying on a 3rd party open source of lens correction data.
    Last edited by Regor250; 12-31-2020, 08:41 AM.

  • #2
    Originally posted by Regor250 View Post
    Just wondering, would it be nice if ACDSee were to implement an in-application lens profiling feature. Right now it relies on a third party lens correction database, which isn't up-to-date and not very helpful if one owns a lens that isn't in the database, but couldn't ACDSee develop an algorith to reverse engineer the in-camera corrections? It could be as simple as providing 2 images per Focal Length, one in-camera JPEG and the RAW version of the sames, and compare the two. It could even use the embedded JPEG preview in the RAW file as the base for comparison. The idea seem so simple that one has to wonder why they aren't doing it already by default instead of relying on a 3rd party open source of lens correction data.
    Interesting.

    Whilst comparing the RAW with an ex-camera JPG might work for fixed lens cameras where the internal firmware in the camera that processes ex-camera JPG's knows the characteristics of the lens, and maybe even for interchangeable lens cameras with their own brand lenses, I'm not so sure it would work for cameras that were being used with 3rd party lenses.

    I believe a significant amount of RAW development software uses for instance LensFun calibration data for their lens correction. I can see one immediate advantage of that. When I read about the method used to arrive at the LensFun calibration, it seems to me that there could be issues if the target images are not ideal, for instance if there are perspective errors in the images, but regardless of whether the end data is 100% correct, every one using a particular software program with the published LensFun calibration data should get the same results as the next person. If it is left to individuals to generate their own lens correction data that would not be the case.

    One the other hand, one has to wonder what will happen, if and when LensFun stops being actively supported?.

    Comment


    • #3
      The use of LensFun correction database makes a lot of sense, but assuming I take all the shots necessary to profile my 12-45/4 lens into the database (which in itself is a lot of work), and someone profiles it, then it gets added to the database (months later), then how do I get that new dataset to work with ACDSee? I suppose that I could also manually make the lens corrections from a calibration chart and save those as presets, but again a lot of work while I can simply develop the RAW in Olympus Workspace, which is what I have to do right now since the lens isn't profiled. A utility to create one's own profile like I suggested would be awesome I think and encourage the use of ACDSee as one's only RAW developing app. Also as you said, what happens when LensFun is no longer supported, or simply far behind? ACDSee lens profiles could be saved and shared with ACDSee Commuity for user download and I think would be attractive to many.
      Last edited by Regor250; 01-01-2021, 11:35 AM.

      Comment


      • #4
        Every statement you make points back to yourself. It's up to you contribute to the lensfun project. You may start with supplying suitable raws or go the extra mile and create the whole xml file yourself. The whole procedure is public and well known. It also is yours to ensure that the project will not dry out.

        Imho, you should be glad that AC uses lensfun and ACDsystems did not try to reinvent the wheel, which would just lead to another square, after all the nonsense they created in the past (eg. collections).

        Comment


        • #5
          Tx Emil for you insightful and helpful contribution. You obviously have profiled several lenses already, so perhaps you can tell us how to update the local Lensfun database from ACDSee install without having to wait for the next update release, then it's just a matter of spending a few hours of effort, which I might if I knew.

          That said, a year ago, LensFun project reported the following: "Currently, I do not have time to maintain the Lensfun project in way that is satisfactory for myself, for the community and for the users. This project really needs someone who takes a look at the bigger picture and has the vision and motivation to keep the project alive and in a good condition. Therefore, I would like to ask the community to take over the Lensfun project management and maintenance in the next months."

          The LensFun Coverage list does not show a single lens in the process of being profiled and even if I were to submit profiling information, there are doubts that the database is still being maintained. So where is that taking us?
          Last edited by Regor250; 01-06-2021, 08:35 AM.

          Comment


          • #6
            Read this and this. You may also want to try replacing the content of your LensFun folder with the content of this package, which just is 12 days old.

            Comment

            Working...
            X